Skip to content

JPeetz/agent-skills / skills/data/data-analysis

data-analysis

Verified

Use this skill when analyzing structured datasets: load, clean, explore, visualize, and report on data from CSV, JSON, Excel, or SQL sources. Produces statistical summaries, correlation matrices, time series analysis, regression models, hypothesis tests, and publication-quality visualizations.

FileKindSize
CHANGELOG.mdmarkdown719 B
LICENSEother1.0 KB
SKILL.mdskill md12.1 KB
evals/eval_cases.jsonconfig6.2 KB
evals/evals.jsonconfig3.3 KB
references/data-cleaning-guide.mdmarkdown8.3 KB
references/statistical-methods.mdmarkdown13.3 KB
references/visualization-patterns.mdmarkdown12.2 KB
scripts/validate_dataset.pyscript20.5 KB

evals/evals.json

View on GitHub
[
  {
    "id": "da-001",
    "description": "Load and inspect a CSV dataset",
    "input": "Analyze this CSV — what patterns do you see in sales.csv?",
    "shouldTrigger": true,
    "expect": "Agent loads the CSV into a DataFrame, checks encoding/delimiter, and reports shape, dtypes, and summary statistics before deeper analysis.",
    "assertions": [
      "dataset is loaded with pd.read_csv and inspected with info()/describe()",
      "shape, columns, dtypes, and numeric summary are reported",
      "loading happens before any analysis"
    ]
  },
  {
    "id": "da-002",
    "description": "Time series trend visualization",
    "input": "Plot revenue over time and show me the trend.",
    "shouldTrigger": true,
    "expect": "Agent sets a datetime index, resamples as needed, and produces a line/area chart with labeled axes, title, and data source footnote.",
    "assertions": [
      "date column is parsed with pd.to_datetime and set as index",
      "a line or area chart is produced for time x numeric",
      "axes are labeled and chart has a title"
    ]
  },
  {
    "id": "da-003",
    "description": "Clean a messy dataset without mutating source",
    "input": "Clean this messy dataset — fix missing values and deduplicate.",
    "shouldTrigger": true,
    "expect": "Agent works on df_clean = df.copy(), reports each change (duplicates dropped, imputation used), and never modifies the source file without permission.",
    "assertions": [
      "work happens on a copy, not the original",
      "each cleaning action is reported with counts",
      "irreversible decisions are surfaced before acting"
    ]
  },
  {
    "id": "da-004",
    "description": "Statistical hypothesis test with honest reporting",
    "input": "Run a t-test on group A vs B to see if the means differ.",
    "shouldTrigger": true,
    "expect": "Agent runs an independent t-test (checking assumptions), and reports test statistic, p-value, and effect size rather than just significant/not.",
    "assertions": [
      "ttest_ind or Mann-Whitney U is used per assumptions",
      "p-value and effect size are reported",
      "sample size caveats are mentioned when relevant"
    ]
  },
  {
    "id": "da-005",
    "description": "Near-miss negative: database schema design without data",
    "input": "What columns should my users table have?",
    "shouldTrigger": false,
    "expect": "Schema design without actual data is an explicit near-miss negative for this skill.",
    "assertions": [
      "skill does not trigger on schema design questions"
    ]
  },
  {
    "id": "da-006",
    "description": "Near-miss negative: statistics theory without a dataset",
    "input": "Explain the central limit theorem.",
    "shouldTrigger": false,
    "expect": "General statistics theory without a dataset context is a documented near-miss negative.",
    "assertions": [
      "skill does not trigger on pure theory questions"
    ]
  },
  {
    "id": "da-007",
    "description": "Near-miss negative: ETL pipeline architecture",
    "input": "Design a data ingestion pipeline for our warehouse.",
    "shouldTrigger": false,
    "expect": "ETL pipeline architecture is a documented near-miss negative — routes to data engineering skills instead.",
    "assertions": [
      "skill does not trigger on pipeline architecture requests"
    ]
  }
]
finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the workStars: 322kUpdated Claude CodeCodexVerifiedInstall
systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixesStars: 322kUpdated Claude CodeCodexVerifiedInstall
test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation codeStars: 322kUpdated Claude CodeCodexVerifiedInstall
skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.Stars: 198kUpdated Claude CodeCodexVerifiedInstall